Phosphorylation of Hdmx mediates its Hdm2- and ATM-dependent degradation in response to DNA damage

Abstract

Maintenance of genomic stability depends on the DNA damage response, an extensive signaling network that is activated by DNA lesions such as double-strand breaks (DSBs). The primary activator of the mammalian DSB response is the nuclear protein kinase ataxia-telangiectasia, mutated (ATM), which phosphorylates key players in various arms of this network. The activation and stabilization of the p53 protein play a major role in the DNA damage response and are mediated by ATM-dependent posttranslational modifications of p53 and Mdm2, a ubiquitin ligase of p53. p53's response to DNA damage also depends on Mdm2-dependent proteolysis of Mdmx, a homologue of Mdm2 that represses p53's transactivation function. Here we show that efficient damage-induced degradation of human Hdmx depends on functional ATM and at least three sites on the Hdmx that are phosphorylated in response to DSBs. One of these sites, S403, is a direct ATM target. Accordingly, each of these sites is important for Hdm2-mediated ubiquitination of Hdmx after DSB induction. These results demonstrate a sophisticated mechanism whereby ATM fine-tunes the optimal activation of p53 by simultaneously modifying each player in the process.

Fig. 1.

ATM-dependent decrease in the cellular amount of Hdmx and its ubiquitination in response to treatment with DSB-inducing agents. (a) Wild-type and A-T (ATM-deficient) lymphoblastoid cells were treated with IR (3 Gy), NCS (100 ng/ml), or etoposide (5 μM). Cells were harvested at the indicated time points, and cellular lysates were subjected to Western blotting analysis. Phosphorylation of p53 on Ser-15, an ATM target site (7, 23), served to monitor the DNA damage response. (b) Ectopic Hdmx responds to DSBs similarly to the endogenous protein. MCF-7 cells were transfected in 6-cm dishes with 200 ng HA-Hdmx. Twenty-four hours after transfection, the cells were treated with etoposide (20 μM), NCS (500 ng/ml), or IR (10 Gy), harvested 5 and 7 h later, and cellular lysates underwent Western blotting analysis. (c) Effect of the ATM-specific inhibitor KU-55933 on DNA damage-induced decrease in cellular amount of Hdmx. MCF-7 cells were pretreated with KU-55933 (10 μM) or DMSO (mock treatment) for 1 h and subsequently with NCS (100 ng/ml) and harvested at the indicated time points. Note the appearance of a slower migrating Hdmx band (band shift) after NCS treatment. Both band shift and Hdmx degradation are eliminated by pretreatment with KU-55933. (d) ATM-dependent ubiquitination of Hdmx in response to DSBs. MCF-7 cells were transiently transfected with HA-Hdmx (500 ng) and Hdm2 (100 ng) expression vectors in combination with CMV-LacZ and CMV-His-6-ubiquitin. MG132 (20 μM) and KU-55933 (10 μM) were added 24 h after transfection as indicated. The proteasome inhibitor MG132 was used to prevent rapid degradation of ubiquitinated Hdmx, thus allowing accumulation of ubiquitinated Hdmx. Thirty minutes later, NCS (200 ng/ml) was added for 3 h, after which the cells were harvested. Lysates were analyzed for ubiquitinated proteins as described in Materials and Methods.(Upper) His-6-purified proteins analyzed for ubiquitinated Hdmx by using anti-HA antibody. Total lysates (Lower) were analyzed for p53 phosphorylation on Ser-15 (DNA damage control), total p53, LacZ (transfection control), HA-Hdmx proteins, and Hdm2 expression. Note the marked suppression of damage-induced Hdmx ubiquitination by the ATM inhibitor.

Fig. 2.

Phosphorylation of Hdmx on S403 in response to DNA damage. (a) The DNA damage-induced band shift of ectopic Hdmx is abolished by a S403A mutation, suggesting that this band shift had been caused by a modification that occurred at that site. U2-OS cells in 6-cm dishes were transfected with 200 ng of wild-type or S403 mutant Hdmx expression vectors and 24 h later were treated with the indicated doses of NCS for 1 h. Hdmx*, modified Hdmx. (b) Ectopic Hdmx is phosphorylated on S403 in response to DNA damage. U2-OS cells in 6-cm dishes were transfected with 1.5 μg of wild-type or S403A mutant Hdmx vectors and treated 24 h later with the indicated doses of NCS for 1 h. The S403 phospho-specific antibody (Fig. 8) detects the phosphorylation of this site on ectopic wild-type Hdmx but not on S403A mutant protein. Because of the large excess of ectopic Hdmx and the different running conditions, the band shift observed in a is not seen in this experiment.

Fig. 3.

ATM dependence of Hdmx phosphorylation on S403. (a) Endogenous Hdmx was immunoprecipitated with anti-Hdmx antibodies (a mixture of the mouse monoclonal antibodies 6B1A, 11F4D, and 12G11G) from lymphoblastoid cells with the indicated genotypes after treatment with 100 ng/ml NCS for 15 min. Western blotting analysis was carried out by using the anti-phospho-Ser-403 antibody. (b) Primary fibroblasts from a healthy donor and an A-T patient were transfected by electroporation with an Hdmx expression vector and treated 48 h later with 200 ng/ml NCS for 1 h. Hdmx was immunoprecipitated with anti-Hdmx antibodies (6B1A, 11F4D, and 12G11G). Immune complexes and cellular lysates were analyzed by Western blotting.

Fig. 4.

Effect of mutating phosphorylation sites on the DNA damage-induced degradation of Hdmx, and stimulation of this process by Hdm2. (a) MCF-7 cells (6-cm dishes) were transiently transfected with 200 ng of vectors expressing wild-type or mutant versions of HA-Hdmx. Twenty-four hours later, the cells were treated with etoposide (20 μM) for 5 h, and total cell lysates were analyzed by Western blotting. Note the moderate reduction in cellular amount of ectopic Hdmx, which is abolished in the mutant proteins. (b) Enhancement of DNA damage-induced reduction in ectopic Hdmx levels by coexpression of Hdm2. MCF-7 cells (6-cm dishes) were transiently transfected with 100 ng of vector expressing Hdm2 and 500 ng of vectors expressing wild-type or S403A HA-Hdmx. Twenty-four hours later, the cells were treated with NCS (500 ng/ml) for 5 h, and total cell lysates were analyzed by Western blotting. Note the stimulation of damage-induced Hdmx degradation by increasing amounts of ectopic Hdm2. Under these conditions, the inhibitory effect of the S403 mutation is evident. (c) Effect of various combinations of mutations on DNA damage-induced reduction of ectopic Hdmx levels. MCF-7 cells (6-cm dishes) were transiently transfected with 500 ng Hdmx expression vectors together with 100 ng of Hdm2 vector. Twenty-four hours later, the cells were treated with NCS (500 ng/ml) for 5 h, and total cell lysates were processed for Western blotting analysis. Of the four phosphorylation sites identified in this study, T365 does not seem to be needed for DNA damage-induced degradation of Hdmx.

Fig. 5.

Effect of mutating phosphorylation sites of Hdmx on its DSB-induced ubiquitination. The same experimental system as in Fig. 1d was used. The MCF-7 cells (6-cm dishes) were transfected with 500 ng of wild-type or mutant HA-Hdmx as indicated and 100 ng of Hdm2 expression vectors, again in the presence of CMV-LacZ and CMV-His-6-Ubiquitin expression vectors. Twenty-four hours after transfection, MG132 (20 μM) was added, and, 15 min later, cells were incubated with NCS (500 ng/ml) for 6 h. The rest of the experiment was carried out as in Fig. 1d. The mutant versions of Hdmx in which one or two phosphorylation sites are mutated demonstrate different degrees of requirement of these sites for Hdmx ubiquitination.